Cost-effective, optimal use of datacenter capacity: Today, tomorrow and beyond

Platform:

Quickly install, deploy & manage

  • The WEB UI drives a solution installer to automatically install and configure the platform components and solution kit modules.
  • A sample database is provided that can be used to verify that all of Capacity Management Analytics is properly installed and configured. It is an optional selection for installation that when selected will create the sample data and then load that data. Subsequent Solution Kit modules will use this data allowing for the verification of the streams and reports for those modules.
  • The installation, configuration, and management of the Solution Kit can also be done via command line instead of through the WEB UI and Solution Installer. This is useful for continued control of solution kit features after the initial installation and configuration. It is also useful in operating systems where the solution installer is not supported.
  • A preview mode is provided for the installation, configuration, and management of the Solution Kit. This allows you to see what will be executed before running it so you can review and make sure you organization’s standards are being followed.
  • The Solution Kit installation is separated into a CMAHOME directory and a CMAINSTANCE directory. The CMAHOME is for executable code, configuration information, and data that should not be changed after the initial installation. The CMAINSTANCE directory is for configuration and data specific to an individual instance of Capacity Management Analytics. The CMAHOME can be used by multiple CMAINSTANCE(s) each with its own directory and Capacity Management Analytics metadata database.

Track, measure and forecast datacenter capacity and usage and analyze software costs

  • Helps manage the lifecycle of your datacenter with a solution that includes capabilities from IBM Cognos® Analytics, SPSS® Modeler, ILOG CPLEX Optimization and a set of prepackaged interactive reports, all of which is integrated with IBM Tivoli® Decision Support for z/OS
  • Compiles the information for the more cost-effective and easier analysis of usage, service objectives, resource utilization, system tuning, accounting and cost recovery and includes software cost analysis to help you manage costs
  • Measures and manages available and planned capacity and balances workload based on business targets
  • Enables you to plan capacity for the future based on historical usage
  • Creates what-if scenarios as activity grows

Jumpstart your capacity management

  • Expedites adoption and value with a set of prepackaged reports that provide insights into processor, LPAR, workload and memory usage and can be expanded to meet changing information needs
  • Includes graphical tools that simplify the creation of more reports and dashboards as needed to help you manage your capacity
  • Offers, as an option, a service that provides experienced IBM specialists who use IBM proven practices to accelerate your solution deployment, along with 40 hours of post-implementation application support and mentoring
  • Provides, as an optional add-on, training, education and mentoring
  • Includes a set of reports to manage capacity for distributed systems and provides a workspace that integrates zEnterprise servers and distributed systems.

Pinpoint issues or potential issues

  • Establish and incorporate monitoring thresholds so that you are alerted to and can address potential issues before they affect performance.
  • Take advantage of a top-down view of zEnterprise workloads with the ability to drill into further detail and identify possible breaking points.
  • Conduct ad hoc analysis to get to the "why" for things such as usage goals that are not achieved, a network path that presents a possible delay or is blocked, and too much or too little storage.
  • Use real-time anomaly detection to uncover issues.
  • View delays that are affecting processing and identify peak utilization periods.

Compare actual usage to forecast and predict future requirements

  • Forecast the capacity needed for such things as supporting end-of-year processing or the launch of a new application.
  • Predict the impact of future sales promotions on critical business applications.
  • Identify which workloads are consuming critical system resources and project their impact over time.
  • Monitor current usage in real-time and compare actual to planned usage so that anomalies are identified before they affect the business.

Communicate datacenter performance information

  • Provide datacenter activity, capacity and usage information to IT and business decision-makers and when, where and how they need it.
  • Extend capacity analytics to mobile devices, including tablet computers, so users can view and fully interact with datacenter information in a security-rich environment.
  • Deliver the foundation and necessary insight needed for all involved to make the best datacenter capacity decisions for the business.

Solution Kit Modules:

Application Analytics

  • Application Analytics reporting on z/OS is enabled by using an application mapping table where Report Classes and task names are mapped to Applications and Functions within Applications. You can use this reporting to focus your attention on which applications use the most MIPS and where they use those MIPS within the environment.
  • Application Analytics reporting on distributed operating systems is enabled by using an application mapping table where servers are identified as serving an Application and Functions within Applications. You can use this reporting to see which applications consume the most servers and which of the servers that are supporting an application are over or under utilized.
  • Application Analytics modeling is enabled by using the Application mapping table to provide the full picture of how many MIPS (regardless of where) an Application has consumed over time. The modeling is done against this Application view of historical overall usage and it allows you to predict the usage into the future.

Software Cost Analysis

  • Software cost analysis summary: Shows MSU utilization and pricing information for all the registered IBM products for CPCs within an enterprise for each scenario, observed, forecasted, and optimized. A listing of IBM NO89, unregistered, products is also provided.
  • LPAR MSU Utilization: Shows various MSU utilization metrics in different scenarios from an LPAR perspective, such as LPAR utilization MSU, LPAR for Hour Rolling Average (HRA) utilization MSU, LPAR billable MSU and product MSU utilization for each scenario, observed (can also display data at the System Management Facility (SMF) collection interval), forecasted, and optimized.
  • Product MSU and Price: Shows the IBM product billable MSUs and monetary cost, from a single CPC, and the LPARs within, perspective for each scenario, observed, forecasted, and optimized.
  • No SMF 89 (NO89) Product MSU and Price: Shows the billable MSUs for the IBM NO89, unregistered, products along with their price.
  • Product optimization: Uses the LPAR and product MSU forecasts to recommend the optimal LPAR and product placement combination to reduce overall MSU usage.

Capacity Planning & Forecasting

  • Peak utilization reporting allows you to display the peak utilization instead of the average, and to show when the peaks happen. When you look at MIPS usage data for a selected interval, you can see the average, or the peak usage over each interval across the selected date range. You can show peak usage by using different calculation methods.
  • Scenario modeling lets you decide how to allocate processing costs to the zIIP/zAAP engines where it is most efficient and cost effective. Operational improvements that are based on the best scenario can drive significant cost reductions and savings.
  • zEnterprise dashboard provides a quick visual status of a zServer, its LPARs, and disk subsystems. Key metrics are provided at the overall machine level. CPU and Storage metrics and I/O and Network metrics are displayed at the LPAR level.

Systems Management & Optimization

  • Enterprise dashboard: Shows high-level information for all the supported servers across the enterprise such as server counts, and percentile of underutilized and overutilized servers.
  • CPU over/under share weight reporting shows the view of the shared processor pool on a CPC with respect to which LPARs are using it and how much they are using. The reporting also shows the usage percentage over or under its guaranteed share of each LPAR due to the competition among the LPARs.
  • LPAR weight optimization uses IBM ILOG® CPEX Optimization Studio to provide recommendations for best weight value for each LPAR within a CPC. The optimization streams use representative dates and time ranges of the typical usage that you want to optimize your environment for, and provides recommended LPAR weights to meet your needs.

Problem Identification

  • Customer Information Control System (CICS®) anomaly detection helps to improve system management by proactively detecting anomalies in forecast capacity so that you can identify issues before they affect the business.

CPU: MIPS Used - IBM System z and LPAR Level

This report shows CPU MIPS used at the System z and central electronic complex (CEC) level and is broken down to the logical partition (LPAR) level. CPU MIPS used can be displayed for each processor type, IBM System z Integrated Information Processor (zIIP), System z Application Assist Processor (zAAP), and Integrated Facility for Linux (IFL).

CPU: MIPS Used - IBM System z and LPAR Level

CPU MIPS Used – System Level (Captured versus Uncaptured)

This report shows captured and uncaptured CPU MIPS used at the system level. The system's capture ratio is also displayed. A notable increase or decrease in a system's capture ratio might indicate a system problem so you should closely monitor a system's capture ratio after any CPU hardware changes and major operating system and software upgrades.

CPU MIPS Used – System Level (Captured versus Uncaptured)

CPU MIPS Used – Service Class Period

This report shows MIPS used at the workload manager (WLM) service class period level. The total CPU time used by all WLM service class periods is known as captured CPU time.

CPU MIPS Used – Service Class Period

CPU: MIPS Used - System z and LPAR Level with Forecast

This report shows actual CPU MIPS used at the System z and CEC level plus forecasted CPU MIPS usage. CPU MIPS used (actual and forecasted) can be displayed for each processor type (CP, zIIP, zAAP, and IFL). Forecasting CPU MIPS usage is critical to the capacity planning process to ensure adequate CPU capacity will be available to support future growth.

CPU: MIPS Used - System z and LPAR Level with Forecast

CPU: Latent demand

This report shows the latent demand (In and Ready count) and capture ratio for a system. When the In and Ready count exceeds the number of logical processors assigned to a system, it creates latent demand, or work that is waiting to be dispatched on a processor because all processors are busy.

CPU: Latent demand

Memory: Common Service Area (CSA), Extended Common Service Area (ECSA), System Queue Area (SQA) and Extended System Queue Area (ESQA) Utilization

This report shows the peak (maximum) utilization for the common virtual CSA, ECSA, SQA and ESQA storage areas. System outages can occur when CSA or ECSA storage is exhausted so monitoring CSA, ECSA, SQA, and ESQA utilization is critical.

Memory: Common Service Area (CSA), Extended Common Service Area (ECSA), System Queue Area (SQA) and Extended System Queue Area (ESQA) Utilization

WLM: Performance Indexes

This report shows the distribution of WLM for eight performance index ranges. It provides a quick visual of how often WLM goals are being met (performance index equal or less than 1) or missed (performance index greater than 1) and to what degree.

WLM: Performance Indexes

WLM: Delays by ImportanceLevel

This report shows the percentage of time that delays (CPU, IO and paging) are negatively affecting work. Delays at the service class period level are aggregated and displayed for each WLM importance level.

WLM: Delays by ImportanceLevel

WLM: Delays by Service Class Period

This report shows the percentage of time when delays (CPU, IO, paging and more) are negatively affecting work. Delays are displayed at the service class period level.

WLM: Delays by Service Class Period

Capacity Management Analytics Screenshots

Capacity Management Analytics resources

IBM Capacity Management Analytics (CMA) turns machine data to insight and provides built-in expertise in capacity management, which empowers IT infrastructure systems managers, system administrators, and data center executives to optimize IT performance, availability, and scalability. CMA brings in the power descriptive, predictive and prescriptive analytics that enables user to more quickly turn machine big data into actionable insight. CMA consists of an operational intelligence platform and expert-based Capacity Management Analytics application modules, the CMA Solution Kit.

The IBM Capacity Management Analytics platform is designed to help you:

  • Quickly install, deploy & manage to gain insight & improve time to value.
  • Track, measure and forecast capacity and usage and analyze software costs with IBM analytics.
  • Jumpstart your capacity management with a set of prebuilt interactive reports that provide information for the entire datacenter, an integrated workspace and optional deployment services and education.
  • Pinpoint issues or potential issues before they adversely affect user experience with capabilities that include real-time anomaly detection.
  • Compare actual usage to forecast in real-time so you can quickly identify and resolve anomalies and predict future requirements.
  • Communicate usage information to various decision-makers in IT and the business.

The CMA Expert-based Application Modules provides built-in expertise on the core capabilities that a data center manager requires today and tomorrow to manage the entire data center, both z Systems and distributed end to end. It consists of a predefined data model to warehouse machine data from across the data center, a set of predefined reports to visualize what is defined and how it is being used, a set of predictive models to understand what is needed in the future and likely usage scenarios, and a set of prescriptive models to optimize the use of IT resources by producing a best possible configuration, action plan, or schedule resources.

The CMA Solution Kit is designed to provide insight into:

  • Application Analytics focus your attention on which applications use the most CPU and where they use those CPU within the environment.
  • Software Cost Analysis assists in managing z/OS® software costs and helps you to identify where and when workloads need to be adjusted and when additional capacity is required.
  • Capacity Planning & Forecasting your future requirements to ensure that capacity is available when the business needs it.
  • Systems Management & Optimization with sophisticated reporting capabilities & optimization mechanisms that analysts, systems managers, and executives can use to view, interact, and personalize their reporting to suit their requirements.
  • Problem Identification improves system management by detecting anomalies.

IBM Capacity Management Analytics is built on IBM Cognos Analytics, IBM SPSS Modeler Gold, IBM ILOG CPLEX Optimization Studio, IBM Solution Installer, IBM Capacity Management Analytics Solution Kit, and IBM Tivoli Decision Support for z/OS.

  • IBM Cognos Analytics provides a complete range of analysis capabilities necessary to optimize the performance and return on investment of z Systems™ and distributed systems. With the industry-leading reporting, analysis, dashboards, collaboration, and mobile business intelligence on a single infrastructure, you can confidently manage the system activity while significantly improving your ability to identify potential issues and pinpoint their cause. You can author, share, and analyze all the data from all the enterprise sources necessary for a complete view of the ever-changing business conditions.
  • IBM SPSS Modeler Gold helps organizations use predictive analytics to forecast their future requirements and ensure the capacity required is available when the business needs it. You can discover the hidden relationships in platform data and predict the outcome of future events and interactions. IBM SPSS Modeler Gold's easy-to-use graphical interface puts the power of data mining in the hands of those responsible for managing the capacity of z Systems and distributed systems, so they can discover new insight and increase productivity. It allows organizations to quickly realize a positive return on investment (ROI).
  • IBM ILOG CPLEX Optimization Studio provides capabilities to optimize the use of IT resources by producing a best possible configuration, action plan, or schedule of resources.
  • Tight integration with IBM Tivoli Decision Support for z/OS enables data collection and provides easy access to that data by Cognos Business Intelligence and SPSS Modeler for all analysis. It ensures that IT can utilize the information necessary for use in performance reporting, service-level management, and usage accounting.

IBM Capacity Management Analytics comes on two platforms, the later platform comes bundled with Tivoli Decision Support for z/OS. All components, other than Tivoli Decision Support for z/OS, can run on distributed platforms like Windows, Linux for System z or Linux for System x.

IBM Capacity Management Analytics

  • CMA Solution Kit
  • IBM Solution Installer
  • IBM Cognos Analytics
  • IBM SPSS Modeler Gold
  • IBM ILOG CPLEX

 

IBM Capacity Management Analytics on z/OS

  • CMA Solution Kit
  • IBM Solution Installer
  • IBM Cognos Analytics
  • IBM SPSS Modeler Gold
  • IBM ILOG CPLEX
  • IBM Tivoli Decision Support for z/OS

 

Not in United States?

Contact IBM

Considering a purchase?